EO for Poverty: Developing Metrics to Support Decision Making Using Earth Observation

It is estimated using official UN statistics that there will be between 570 and 596 million people living in extreme poverty in 2030 and thus the UN Sustainable Development Goal 1 “eradicate poverty in all its form everywhere” will not be achieved. Measurement of socioeconomic conditions underpins the global goals and huge amounts of work have been dedicated to ensuring all 17 goals, and the 169 targets, have associated metrics that can be used to monitor progress. Despite these advances, a shortage of frequent, high-quality data still hinders the measurement of socioeconomic conditions such as poverty. The lack of regularly updated poverty metrics prevents timely policy adjustments. There are a little over six years remaining of the SDG timeframe (2015–2030), little time in which to begin collecting more poverty data in traditional ways. Over recent years there has been a steady increase in interest around the use of Earth Observation satellite imagery to predict poverty. The reasons for doings so are often given as plugging various data gaps to support decision makers in monitoring progress towards the global goals. However, some of the prevailing methods used do not lend themselves well to decision making and policy. This chapter introduces the background to the sub-field of EO for poverty, provides some details on the methods used to date to do this and then critiques the current approaches. The focus of these critiques is on how the data can be used to support poverty alleviation/anti-poverty policy creation.